The Role of Hybridization in the Evolution of Reef Corals

ANRV292-ES37-18 ARI 17 October 2006 7:29

Bette L. Willis,1 Madeleine J.H. van Oppen,2 David J. Miller,3 Steve V. Vollmer,4 and David J. Ayre5

1Centre of Excellence for Coral Reef Studies, School of Marine Biology and Aquaculture, James Cook University, Townsville, Queensland 4811, Australia; email: [email protected] 2Australian Institute of Marine Science, Townsville MC, Queensland 4810, Australia; email: [email protected] 3Comparative Genomics Center, James Cook University, Townsville, Queensland 4811, Australia; email: [email protected] 4Smithsonian Tropical Research Institute, Balboa, Panama: email: [email protected] 5School of Biological Sciences, University of Wollongong, New South Wales 2522, Australia; email: [email protected]

Annu. Rev. Ecol. Evol. Syst. 2006. 37:489–517 Key Words First published online as a Review in Advance Acropora, hybrid ﬁtness, introgression, reticulate evolution, sperm on August 17, 2006 choice crosses The Annual Review of Ecology, Evolution, and Systematics is online at Abstract http://ecolsys.annualreviews.org The importance of hybridization in the evolution of plant species This article’s doi: 10.1146/annurev.ecolsys.37.091305.110136 is widely accepted, but its contributions to animal species evolution remain less recognized. Here we review evidence that hybridization Copyright c 2006 by Annual Reviews. All rights reserved has contributed to the evolution of reef corals, a group underpin- ning the coral reef ecosystem. Increasingly threatened by human 1543-592X/06/1201-0489$20.00

by NORTHEASTERN UNIVERSITY on 01/11/07. For personal use only. and climate-related impacts, there is need to understand the evolutionary processes that have given rise to their diversity and contribute to their resilience. Reticulate evolutionary pathways among

Annu. Rev. Ecol. Evol. Syst. 2006.37:489-517. Downloaded from arjournals.annualreviews.org the ecologically prominent, mass-spawning genus Acropora suggest that hybridization, although rare on ecological timescales, has been instrumental in their diversiﬁcation on evolutionary timescales. Ev- idence that coral hybrids colonize marginal habitats distinct from those of parental species’ and that hybridization may be more frequent at peripheral boundaries of species’ ranges supports a role for hybridization in range expansion and adaptation to changing environments. We conclude that outcomes of hybridization are signiﬁ- cant for the future resilience of reef corals and warrant inclusion in conservation strategies.

489 ANRV292-ES37-18 ARI 17 October 2006 7:29

INTRODUCTION Hybridization has long been considered a major creative force in the evolution of plant species (e.g., Anderson 1949, Anderson & Stebbins 1954, Grant 1981, Stebbins 1959), but traditionally, its role in the evolution of animal species has been thought to be inconsequential, or at most, significant only in terms of reinforcing barriers to interspecific fertilization (Dobzhansky 1937, Mayr 1963). However, as predicted in a comparatively recent review of animal hybridization (Dowling & Secor 1997), evidence is mounting that hybridization is more common among animal species than previously thought and occurs in the majority of phyla in both terrestrial and marine habitats (reviewed in Arnold 1997, Gardner 1997, Mallet 2005). Phylogenetic hotspots, where (as in British vascular plants) approximately 25% of animal species hybridize, include British duck and game bird species, and American warblers and butterflies, but overall, 6–12% of bird, butterfly, and mammal species hybridize natu- rally (Mallet 2005). There are comparatively few studies of hybridization in modular, sessile marine animals (Gardner 1997), whose modes of speciation are more likely to resemble those of plants given similarities in their life histories. For example, both have sessile modular adults, frequent hermaphroditism, absence of mating behavior, broadcast spawning of gametes, passive dispersal of juveniles, and extensive asexual reproduction. Further studies of animals with plant-like life histories will enhance current understanding of the evolutionary significance of hybridization in animals. On tropical coral reefs, the simultaneous mass spawning of many species of stony corals (Harrison et al. 1984) represents a unique breeding strategy among animals and suggests that hybridization might have played a role in the evolution of this functionally important group, the cornerstone of the coral reef ecosystem. In the past two decades, an upsurge of studies on the reproduction of scleractinian corals has shown that synchronized spawning among more than two species (i.e., mass spawning sensu Willis et al. 1985) occurs in the majority of reef regions (e.g., Baird et al. 2002, Carroll et al. 2006, de Graaf et al. 1999, Gittings et al. 1992, Guest et al. 2002, Hayashibara et al. 1993, Simpson 1991). In highly synchronized events, up to 35 species in sympatry may spawn within two hours of each other (e.g., mass spawning of corals on Australia’s Great Barrier Reef (GBR); Babcock et al. 1986, Willis et al. 1985). In combination, the apparent absence of temporal barriers to

by NORTHEASTERN UNIVERSITY on 01/11/07. For personal use only. interspeciﬁc breeding provided by mass spawning events, the co-occurrence of large numbers of coral species on reefs where currents mix positively buoyant gametes in a thin layer at the sea surface, and the reliance of the mate recognition system

Annu. Rev. Ecol. Evol. Syst. 2006.37:489-517. Downloaded from arjournals.annualreviews.org on interactions among gametes for assortative fertilization (Palumbi 1994) provide outstanding opportunities for hybridization. Concurrence in the timing of spawning among many coral species and the global nature of the mass-spawning phenomenon suggest that the Scleractinia provide a novel system for gaining insights into the role that hybridization has played in the evolution of animal species. Understanding the evolutionary processes that have shaped modern reef corals and given rise to their diversity has become urgent as the number and intensity of threats to the biodiversity and resilience of coral reef ecosystems escalate against a backdrop of global climate change (e.g., Harvell et al. 2002, Hoegh-Guldberg 1999, Hughes

490 Willis et al. ANRV292-ES37-18 ARI 17 October 2006 7:29

et al. 2003). Estimates that recent declines in coral cover and colony numbers have been greater than 97% for two of the three Atlantic species of the once-common coral genus Acropora and predictions that the downward trajectory may continue (Acropora Biological Review Team2005) have led to their listing as threatened and endangered wildlife under the U.S. Endangered Species Act (ESA). However, evidence that the third coral, Acropora prolifera, is a hybrid and thus fails to meet the ESA definition of a species disqualifies it from consideration for protection under the act. If appropriate management strategies are to be developed for the Scleractinia, greater understanding of the evolutionary processes giving rise to their biodiversity is needed, particularly in relation to the role that hybridization has played in adaptive radiation and evolution within the group. The recent publication of a body of reproductive, morphological, ecological, and molecular data for mass-spawning corals in both the Indo-Pacific and Caribbean (see references in sections below) provides new perspectives on coral evolution and has prompted the present review of the evolutionary importance of hybridization within the Scleractinia.

HYBRIDIZATION AS AN EVOLUTIONARY PROCESS Introgressive hybridization, here defined as the exchange of genes between genetically differentiated species, may have a variety of contrasting outcomes. Extensive and prolonged episodes of introgressive hybridization may lead to the merging of species and ultimately to the extinction of pure parental species along with their morphological, behavioral and/or ecological distinctions. This process of genetic mixing may provide increased genetic diversity, new traits, and heterosis for the emerging species, but may carry the cost of a net loss of species diversity. Regardless of contrasting perspectives on the fate of species following merging and homogenization of gene pools (i.e., the loss of current species versus the emergence of new ones), the comparative speed with which new traits and species can arise suggests that, in some cases, there may be positive selective pressure for hybridization and introgressive events [e.g., among diversifying species of sunflowers (Ungerer et al. 1998) and cyprinid fish (Rosenfield & Kodric-Brown 2003)]. Hybridization may also give rise to new species that are reproductively isolated

by NORTHEASTERN UNIVERSITY on 01/11/07. For personal use only. from parental species through either polyploid speciation or recombinational speciation, both processes representing mechanisms for the rapid evolution of genetic novelty without the loss of parental species. Polyploid speciation involves the produc-

Annu. Rev. Ecol. Evol. Syst. 2006.37:489-517. Downloaded from arjournals.annualreviews.org tion of allopolyploids, which by deﬁnition contain three or more sets of chromosomes from two different species and thus achieve immediate reproductive isolation from parental species. Allopolyploid species are common in plants (Grant 1981, Ramsey & Schemske 1998, Soltis & Soltis 1999) and account for 2–4% of angiosperm species and 7% of fern species (Otto & Whitton 2000). In studies so far, animal allopolyploids are much rarer and are predominantly parthenogenic rather than sexually reproducing (Bullini 1994, Dowling & Secor 1997, Otto & Whitton 2000, White 1978). However, because uniparental species may contribute sexual propagules capable of introgress- ing with other species (e.g., diploid, hemiclonal species in the hybridogenetic ﬁsh

www.annualreviews.org • Hybridization in Corals 491 ANRV292-ES37-18 ARI 17 October 2006 7:29

Poeciliopsis; Mateos & Vrijenhoek 2002) and may have greater distributional ranges than their bisexual ancestors (e.g., Bacillus stick insects, Bullini 1994), lack of sexual viability is not grounds for dismissing the evolutionary potential of asexual hybrid lineages. Recombinational species (Grant 1981), also called homoploid hybrid species, are hybrid species that have the same number of chromosomes as their parent species but are reproductively isolated from them and are true breeding (Grant 1981, Rieseberg 1997). The number of clear-cut cases of recombinational speciation identified suggests that it may be rare in both plants and animals (Coyne & Orr 2004, Rieseberg 1997). For example, Rieseberg (1997) identified only eight well-documented cases of recombinational speciation in plants, three of which are Helianthus sunflower species. However, by its very nature, recombinational speciation, which requires that fit, in- terfertile recombinant hybrid genotypes be produced that are reproductively isolated from their sympatric parent species, is difficult to identify (Rieseberg 1997, Coyne & Orr 2004). The discovery of increasing numbers of diploid hybrid species with the advent of improved molecular techniques suggests that the real number of cases may be much higher than generally thought (Seehausen 2004). Seehausen argues that hybridization is particularly significant where ecological conditions favor adaptive radiations, such as where there are underutilized niches following disturbance or at the ecological or geographical peripheries of species ranges, and based on the preva- lence of hybridization in studies of more recent adaptive radiations [e.g., Darwin’s finches (Freeland & Boag 1999), African cichlids (Salzburger et al. 2002), Hawaiian crickets (Shaw 2002)], its contributions to past species diversifications have been underestimated. Evidence of transgressive segregation, in which phenotypic variation in hybrid populations is greater than the combined variation of parent populations (e.g., enhanced salt tolerance of hybrid Helianthus sunflowers through acquisition of different parental loci with additive effects; Lexer et al. 2003), provides a mechanism by which recombinational species can have greater fitness than parental species. When recombinational speciation does occur, empirical and theoretical evidence suggests that it does so rapidly and is often accompanied by ecological divergence that promotes reproductive isolation (Buerkle et al. 2000, McCarthy et al. 1995, Rieseberg et al. 2003). This further impedes the detection of past hybridization events in the origin and transfer of novel adaptations. Although hybrid zones provide valuable mi- by NORTHEASTERN UNIVERSITY on 01/11/07. For personal use only. crocosms in which current, ongoing evolutionary processes can be studied (Hewitt 2001), the lack of a similar window on past hybridization events makes more distant contributions of hybridization events difficult to detect. Annu. Rev. Ecol. Evol. Syst. 2006.37:489-517. Downloaded from arjournals.annualreviews.org There is wide acceptance among botanists that the transfer of genes between taxa through introgressive hybridization potentially provides more raw material for evolution than can be produced directly by mutation (Anderson 1949, Anderson & Stebbins 1954, Arnold 1997), and hybridization is estimated to have produced a significant fraction (∼11%) of flowering plant lineages (Rieseberg 1997). Even though cases of hybridization in animals continue to be described, the prevalent view is that hybridization has not played a major creative role in the evolution of animal species. This may be partly due to difficulties in identifying the products of hybrid speciation in past events. Even in contemporary studies, hybrid taxa or lineages may

492 Willis et al. ANRV292-ES37-18 ARI 17 October 2006 7:29

be difﬁcult to detect because they may be much closer genetically or morphologically to one parental species than the other ( Jiggins & Mallet 2000, Naisbit et al. 2003). Moreover, establishing that hybrid lineages constitute distinct evolutionary entities that are reproductively isolated from parent species requires multiple lines of evidence including morphological, reproductive, molecular, and chromosomal karyotype data. Without such evidence of effective reproductive isolation, hybrid lineages are hybrids and not hybrid species. In summary, the variety of ways in which hybridization may contribute to the evolution of lineages attests to its scope as an evolutionary process, but evidence of its past contributions may be elusive.

MATING SYSTEMS OF MASS-SPAWNING CORALS The mating systems of corals are based on either broadcast spawning of gametes for external fertilization or internal fertilization followed by larval brooding, with approximately two thirds of species (n = 227) utilizing the former, external mode of development (Harrison & Wallace 1990). The breeding periods of broadcast- spawning corals may be temporally isolated on scales that range from hours (Fukami et al. 2003, Knowlton et al. 1997, Levitan et al. 2004, van Oppen et al. 2002) to weeks or months (Wolstenholme 2004; B.L. Willis, unpublished data), but most corals spawn in tightly synchronized breeding events that overlap with those of a number of species, most commonly congeners (Harrison & Wallace 1990, Richmond & Hunter 1990). Opportunities for interspecific breeding in the mating systems of corals thus vary in response to breeding times that range from complete temporal isolation to complete overlap, with the dominant breeding strategy of synchronized mass spawning providing unparalleled potential for hybridization. Interpreting this potential for hybridization, however, requires knowledge of both prezygotic (including fine-scale temporal barriers and gamete incompatibility) and postzygotic isolating barriers. Next we review isolating barriers for the dominant reef-building genera in two of the major ocean basins: the Acropora in the Indo-Pacific and the Montastraea in the Western Atlantic. The Indo-Pacific Acropora constitute a highly diverse group of more than 100 species, up to 76 of which can occur in sympatry (Wallace 1999; Figure 1a) and at least 35 of which are known to participate in synchronized mass- spawning events (Babcock et al. 1986, Willis et al. 1985). This genus has undergone by NORTHEASTERN UNIVERSITY on 01/11/07. For personal use only. the greatest adaptive radiation of any scleractinian genus in the Indo-Pacific, where it commonly dominates coral communities. In contrast, there are only 3 extant species of Acropora in the Western Atlantic (Figure 1b), indicating that different processes Annu. Rev. Ecol. Evol. Syst. 2006.37:489-517. Downloaded from arjournals.annualreviews.org have shaped the evolution of this genus in these two biogeographic regions.

Prezygotic Isolating Barriers: Mating Precedence in Sperm Choice Experiments Tests of gamete compatibility in laboratory crosses between mass-spawning species have had variable outcomes. Although the data are as yet fragmentary, there are some basic differences in interspeciﬁc gamete compatibility between those Indo- Paciﬁc and Caribbean corals for which data are available. Among species of Acropora

www.annualreviews.org • Hybridization in Corals 493 ANRV292-ES37-18 ARI 17 October 2006 7:29 by NORTHEASTERN UNIVERSITY on 01/11/07. For personal use only. Annu. Rev. Ecol. Evol. Syst. 2006.37:489-517. Downloaded from arjournals.annualreviews.org

Figure 1 Comparison of Acropora (Order: Scleractinia) assemblages on (a) Indo-Paciﬁc reefs, where the genus attains the greatest diversity (more than 100 species) of all extant corals and typically dominates coral communities (almost all corals in the assemblage pictured are species of Acropora); and (b) Caribbean reefs, where there are only 3 extant species: A. cervicornis (top right); A. palmata (bottom right); and the hybrid A. prolifera (bottom left).

494 Willis et al. ANRV292-ES37-18 ARI 17 October 2006 7:29

from the central GBR, outcomes of colony crosses between 38 pairs of species have spanned the complete spectrum of fertilization success from high (50–100%) between 8 pairs, through moderate (10–50%) between 7 pairs, to low (3–10%) between a further 3 pairs of species (van Oppen et al. 2002, Willis et al. 1997). Overall, one third of species pairs of Indo-Pacific Acropora crossed experimentally (n = 73 species combinations) have resulted in greater than 10% fertilization in some colony pairings, based on studies combined from the central GBR (van Oppen et al. 2002, Willis et al. 1997), northern GBR (Wolstenholme 2004), and Japan (Fukami et al. 2003, Hatta et al. 1999). The fraction rises to more than 45% if species pairs with low gamete compatibility (3–10% interspecific fertilization success) in at least some colony crosses are included. Thus the capacity to hybridize appears to be a common feature of mating systems in the Indo-Pacific Acropora, and this feature is consistent over large geographic areas. Similarly, breeding trials between seven morphospecies in the Indo- Pacific genus Platygyra have shown high levels of interspecific fertilization on the GBR (Miller & Babcock 1997, Willis et al. 1997). The lack of genetic distinctiveness among morphospecies of Platygyra in allozyme studies (Miller & Benzie 1997) suggests that hybridization may be more common in this genus than in the Indo-Pacific Acropora. Although laboratory crosses in which eggs are provided only with the opportunity to mate with heterospecific sperm demonstrate that gametes are compatible, significant variation in the relative frequency of inter- and intraspecific fertilization success and dramatic variation among individual pairs of colonies (from 0% to 100% in some species crosses; Fukami et al. 2003, van Oppen et al. 2002, Willis et al. 1997) imply that prezygotic isolating barriers may exist among the Acropora. In order to test for evidence of premating isolation under conditions that more closely resemble in situ mass-spawning events, we carried out a series of sperm-choice breeding trials. We incubated eggs with both conspecific and heterospecific sperm in lab crosses, where sperm came from parents (Acropora millepora and A. pulchra) that were alternately homozygous for allozyme genotypes. Some 14 different sperm-choice crosses were performed, involving 3 cases in which A. millepora was the mother and 11 cases in which A. pulchra was the mother. Electrophoretic analyses of individual larval offspring demonstrated that, in the 11 trials in which A. pulchra was the mother, all offspring (n = 475 larvae) tested were the result of intraspecific fertilizations. Simi- larly, in 2 of 3 sperm-choice trials involving A. millepora as the mother, all offspring by NORTHEASTERN UNIVERSITY on 01/11/07. For personal use only. (n = 117 larvae) tested were purebreds. However, in one trial where A. millepora was the mother, all offspring (n = 120) tested were A. millepora × A.pulchra hybrids. It is conceivable that the colony used as the source of intraspecific sperm in this latter trial Annu. Rev. Ecol. Evol. Syst. 2006.37:489-517. Downloaded from arjournals.annualreviews.org (which resulted in hybrid offspring) had been misidentified and was actually A. spathulata, a cryptic species that has recently been separated from A. millepora (Wallace 1999) on the basis of incompatibility in breeding trials (Willis et al. 1997). However, either hybridization was favored over conspecific fertilization (assuming that the fathers were correctly identified as A. millepora and A. pulchra) or, given a choice between sperm from two different species (A. spathulata and A. pulchra), A. millepora eggs were fertilized by the species with the more compatible sperm. These results suggest that, although eggs of many Indo-Pacific species of Acropora are compatible with heterospecific sperm in no-choice lab crosses, prezygotic isolating mechanisms exist that favor

www.annualreviews.org • Hybridization in Corals 495 ANRV292-ES37-18 ARI 17 October 2006 7:29

conspecific mating in these species at least most of the time. Species specificity in gamete recognition and binding proteins has been found in other marine invertebrates, for example, in abalone (Vacquier et al. 1990) and sea urchins (Metz & Palumbi 1996) and would be a fruitful area of research to further understand the mating systems of mass-spawning corals. Because many species of Acropora are able to hybridize in no-choice experimental crosses, the lack of hybrid offspring produced in the presence of conspecific sperm in the few sperm-choice trials so far performed suggests that prezygotic isolating barriers are semipermeable based on a hierarchy in gamete compatibility. Decreasing gamete compatibility, in rank order from intraspecific to closely related interspecific, and lastly to more distantly related interspecific matings, offers an alternative av- enue for reproductive success if mates are scarce, such as in marginal or disturbed environments. At the same time, precedence of conspecific mating limits gene flow between species, thus maintaining species boundaries and morphologically recognizable species. Regardless of whether the capacity for hybridization under these circumstances is an artifact of their mating systems or a selected feature, the net result is a form of bet-hedging. Although interspecific pairings can lead to gamete wastage, there is ample evidence that they may also result in offspring that, if not sexually viable, are nevertheless able to found clonal lineages that may be ecologically successful (Bullini 1994, Kearney 2005, Spolsky et al. 1992). In the Caribbean, there are no published studies of interspecific gamete compatibility for the genus Acropora, but studies of the Montastraea species complex have found generally low interspecific gamete compatibility between species that have overlapping spawning times. In Panama, evidence of gamete incompatibility in laboratory crosses provides corroborative evidence that M. faveolata, previously thought to be a morphological variant of M. annularis, is a cryptic species (Knowlton et al. 1997, Levitan et al. 2004). A second cryptic species (M. franksi) had moderate (∼40–60%) gamete compatibility with M. annularis, but gametes age rapidly so that fine-scale (two hours) temporal isolation in spawning times provides a reasonably effective reproductive barrier. Significantly, the two species that have the more compatible gametes and are most closely related genetically (i.e., M. franksi and M. annularis) show strong temporal isolation, whereas the two species that spawn together (M. annularis and M. faveolata) have incompatible gametes and are genetically most different (Fukami by NORTHEASTERN UNIVERSITY on 01/11/07. For personal use only. et al. 2004, Levitan et al. 2004). Thus a suite of isolating barriers maintains species boundaries in this group in Panama. Contrasting patterns of gamete compatibility were found at a site near the north- Annu. Rev. Ecol. Evol. Syst. 2006.37:489-517. Downloaded from arjournals.annualreviews.org ern geographical periphery of the ranges of these western Atlantic species. In the Florida Keys, M. faveolata hybridized well with both M. franksi and M. annularis (Szmant et al. 1997). Genetic differences among these three taxa were also weaker in the Bahamas than in Panama (Fukami et al. 2004), potentially reflecting the occasional moderate to high numbers of viable larvae found in some experimental crosses, particularly with aged eggs, between pairs of these three species (Fukami et al. 2004). Similarly, consistent differences in patterns of corallite morphology were detected among these three species across their distributional ranges. Distinct corallite morphologies among all three species in Panama, in contrast to overlapping corallite

496 Willis et al. ANRV292-ES37-18 ARI 17 October 2006 7:29

morphologies in the Bahamas, led Fukami et al. (2004) to speculate that a gradient in hybridization exists; the strongest introgression occurred at northern sites. Differ- ences in gamete compatibility and morphologies among the same species at different geographic locations provide support for Veron’s (1995) suggestion that species may interact differently through hybridization across their ranges. Morphological analyses of fossil specimens from Panama and the Bahamas suggest that present patterns of greatest introgressive hybridization and least morphological differentiation among Montastraea species in the Bahamas have persisted throughout the geological history of this species complex (Budd & Pandolfi 2004). Such persistent differences in the extent of interspecific breeding over broad geographic ranges throughout evolutionary histories are consistent with the notion that semipermeable barriers at the extremities of species’ ranges afford increased opportunities for mating. Patterns of temporal barriers to interbreeding in relation to species relatedness within the Caribbean Montastraea are the inverse of those within the Indo-Pacific Acropora; temporal isolating barriers are strongest between the most closely related species of Caribbean Montastraea (i.e., M. franksi and M. annularis; Levitan et al. 2004) but weakest between closely related species of Indo-Pacific Acropora, where temporal barriers reflect the greatest genetic divergence among species (Fukami et al. 2003; van Oppen et al. 2001, 2002; Wolstenholme 2004). Similarly, patterns of synchronous spawning with gamete compatibility differ between the two reef regions; synchronous spawning in Panama occurs between Montastraea species with the least compatible gametes (Levitan et al. 2004), whereas many Indo-Pacific Acropora and Platygyra species that spawn synchronously have compatible gametes (Miller & Babcock 1997, Fukami et al. 2003, Willis et al. 1997). Gametes of at least some Indo-Pacific species of Platygyra and Acropora remain competent for a significantly longer period of time (6–8 hours post spawning; Miller & Babcock 1997, Willis et al. 1997) than do those of their Caribbean counterparts (1–2 hours, Levitan et al. 2004), contributing to differences between the two regions in the degree of temporal isolation among species that spawn on the same night.

Postzygotic Isolating Barriers: Developmental Competence of Coral Hybrids by NORTHEASTERN UNIVERSITY on 01/11/07. For personal use only. Studies of postzygotic isolating barriers among coral species are scarce, largely because of their long pre-reproductive periods (Harrison & Wallace 1990). In the Indo- Pacific, hybrid offspring from two different species crosses (Acropora millepora × A. Annu. Rev. Ecol. Evol. Syst. 2006.37:489-517. Downloaded from arjournals.annualreviews.org pulchra and A. hyacinthus × A. cytherea) are developmentally competent and have been successfully reared for three years (Figure 2). Moreover, in a large-scale grow-out program on Orpheus Island reefs in the central GBR, patterns in the survival and growth of three-month-old hybrids of A. millepora × A. pulchra in three habitats (i.e., the intertidal, inner reef-flat habitat of A. pulchra, the reef-slope habitat of A. millepora, and an intermediate reef-crest habitat) indicate that, at this early life history stage, their fitness does not differ from that of purebred offspring reared in the reef-flat and reef-crest habitats (Figure 3). Generally, purebred offspring survived best in the habitat where the corresponding parental species was most abundant. Thus offspring

www.annualreviews.org • Hybridization in Corals 497 ANRV292-ES37-18 ARI 17 October 2006 7:29 by NORTHEASTERN UNIVERSITY on 01/11/07. For personal use only. Annu. Rev. Ecol. Evol. Syst. 2006.37:489-517. Downloaded from arjournals.annualreviews.org

Figure 2 Three-year-old coral offspring produced from experimental crosses demonstrate that coral hybrids are developmentally competent. (a) Acropora millepora purebred juvenile; (b) A. pulchra purebred juvenile; (c) A. pulchra (mother) × A. millepora (father) hybrid juvenile; (d) adult A. millepora showing typical corymbose (pillow) morphology and subtidal, reef-slope habitat; and (e) adult A. pulchra showing typical arborescent (branching) morphology and intertidal, reef-ﬂat habitat. Offspring were reared in ﬂow-through tanks at Orpheus Island Research Station, Central Great Barrier Reef, but died in a bleaching event before reproductive competency could be assessed.

498 Willis et al. ANRV292-ES37-18 ARI 17 October 2006 7:29

Figure 3 a 6 10 Hybrids A. pulchra Comparative mean growth A. millepora ± 5 21 SE (a) and survival (b)at3 months among purebred 4 and hybrid offspring of Acropora millepora and A. 74 96 pulchra produced in 3 21 13 98 experimental crosses and outplanted to three habitats 2 (reef ﬂat, crest, and slope) 29 on northeast Orpheus 1 Island reef, central Great Barrier Reef (see text for 0 details of parental habitats). Mean # Polyps / Colony (±SE) / Colony Mean # Polyps Reef flat Reef crest Reef slope Numbers above histograms represent sample sizes: n = # offspring at census 2 (growth); n = # offspring at b 20 504 time 0 (survival).

639 15 104 164 171 605 10 123 % Survival 5 207

133 0 Reef flat Reef crest Reef slope

of A. pulchra survived best on the inner reef flat and reef crest, which are the habitats to which it is largely restricted as an adult. Moreover, A. pulchra offspring grew fastest on the inner reef flat (Figure 3), where the species is most abundant and typically forms monospecific stands. There was a trend toward poorest survival of A. millepora offspring in the intertidal reef-flat habitat, where adults are typically not found, al- by NORTHEASTERN UNIVERSITY on 01/11/07. For personal use only. though their growth in the first three months was not markedly different among the three habitats. Interestingly, patterns of hybrid offspring survival and growth were most similar to those of A. pulchra. Hybrids grew faster than A. millepora on the inner Annu. Rev. Ecol. Evol. Syst. 2006.37:489-517. Downloaded from arjournals.annualreviews.org reef flat (Figure 3a) and also survived better there and on the reef crest than on the

reef slope. Although results for F2-generation hybrids are unknown, results for F1 hybrids are consistent with a review of hybrid fitness by Arnold & Hodges (1995), who found that existing evidence does not support a general pattern of reduced hybrid fitness and concluded that the role of hybridization in the process of evolutionary diversification has been underestimated. The greater growth and survival of hybrids in the more environmentally variable reef-flat and reef-crest habitats suggest a potential role for hybrids in providing a source of variation for adaptation to new or extreme environments.

www.annualreviews.org • Hybridization in Corals 499 ANRV292-ES37-18 ARI 17 October 2006 7:29

In summary, there are few apparent pre- or postzygotic isolating mechanisms that would preclude hybridization from contributing to the evolutionary diversification of at least some of the Indo-Pacific mass-spawning Acropora, particularly when mating opportunities with conspecifics are scarce. However, parallel studies suggest that hybridization has played a different role in the evolution of the dominant Caribbean coral genera. Within the Caribbean Acropora, it has been suggested that postzygotic selection against hybrid genotypes (reviewed below) may be acting as a strong filter against genetic mixing and hence the potential for further evolutionary diversification among the three Caribbean species (Vollmer & Palumbi 2002). Similarly, a variety of mechanisms acting in concert are thought to render hybridization among the three species in the Montastraea complex unlikely in Panama (Levitan et al. 2004), although it appears to have played a greater role in the Bahamas (Fukami et al. 2004). Studies of the fitness of hybrid offspring would shed further light on the nature of postzygotic isolating mechanisms within these two Caribbean species complexes.

MOLECULAR EVIDENCE OF HYBRIDIZATION IN CORALS Hybrid Origin of Coral Species in the Caribbean Acropora: A. prolifera Case Study The only scleractinian coral known to be of hybrid origin, so far, is the Caribbean species Acropora prolifera. Three species of Acropora are known from the Caribbean, all of which are endemic: A. cervicornis, A. palmata and A. prolifera [collectively re- ferred to as the A. cervicornis group (Wallace 1999)]. A. cervicornis and A. palmata are sister species with good fossil records, the earliest fossils being approximately 6.6 (Budd & Johnson 1999) and 3.6–2.6 (McNeill et al. 1997) Myr old respectively. In contrast, A. prolifera is of recent (Holocene) origin and has no fossil record (Budd et al. 1994). The three species are differentiated by colony growth form (Figure 1b) and habitat preference (Cairns 1982, Rutzler¨ & Macintyre 1982). Acropora cervicornis has an arborescent, staghorn morphology and occurs in comparatively sheltered fore- and back-reef habitats. Acropora palmata has a robust, elkhorn morphology and occurs primarily in more exposed, reef-crest habitats. The third described species, Acropora prolifera, has a fused-branched morphology that is intermediate between the by NORTHEASTERN UNIVERSITY on 01/11/07. For personal use only. two species, and has been further differentiated into co-occurring palmate and bushy morphs at one location (in Puerto Rico, Vollmer & Palumbi 2002). It is rarer than the other two Caribbean Acropora species and tends to occur in marginal, shallow-water, Annu. Rev. Ecol. Evol. Syst. 2006.37:489-517. Downloaded from arjournals.annualreviews.org back-reef and reef-crest habitats that the other species do not occupy. All three species are sympatric throughout the Caribbean (Adey et al. 1977, Budd et al. 1994, Goreau 1959, Rutzler¨ & Macintyre 1982, Wallace 1999) and, although they generally occupy different reef zones, their depth distributions may overlap (Adey et al. 1977, Goreau 1959). Despite their distinct morphologies and ecological niches, the taxonomic status of these three species has been debated for over a century (Cairns 1982, Gregory 1895, Vaughan 1901, Wells 1973). The intermediate morphology of A. prolifera and the restriction of its distribution to reefs on which both A. palmata and A. cervicornis co-occur suggest that it may be a hybrid of the other two species.

500 Willis et al. ANRV292-ES37-18 ARI 17 October 2006 7:29

The putative parent species, Acropora palmata and A. cervicornis, spawn synchronously once or twice a year in August and/or September (de Graaf et al. 1999, Szmant 1986; S.V. Vollmer, personal observation) and have compatible gametes in experimental crosses (B.L. Willis, unpublished data); thus opportunities for interspecific hybridization exist. Genetic data confirm that A. prolifera has indeed had a hybrid origin (van Oppen et al. 2000, Vollmer & Palumbi 2002). Allele frequencies of a nuclear intron (PaxC 46/47 intron) were significantly different between sympatric colonies of the parental species; A. cervicornis carried two alleles that were not present in A. palmata at a frequency of 0.769 and 0.039, respectively, confirming that they are distinct species (van Oppen et al. 2000). All colonies of A. prolifera were heterozygous for this locus, the expected outcome if two species with very distinct allele comple- ments and frequencies (i.e., A. palmata and A. cervicornis) hybridize. Ribosomal DNA ITS sequence types were shared among all three species, confirming that A. prolifera is a hybrid between the other two species that backcrosses at low frequency. At two additional nuclear loci, all A. prolifera were also heterozygous for both species’ alleles, indicating that A. prolifera colonies are most likely first generation (F1) hybrids (Vollmer& Palumbi 2002, 2004). Moreover, maternally inherited mitochondrial DNA demonstrated that hybridization occurs in both directions (Vollmer & Palumbi 2002). Three mtDNA haplotypes were observed (A, B, and C), with haplotypes A and C occurring only in A. cervicornis and A. prolifera, whereas haplotype B occurs in all three species. Using phylogenies and Bayesian coalescent models, Vollmer & Palumbi (2002) demonstrated that backcrossing occurs with only one of the parental species, A. cervicornis, thereby passing genes from A. palmata to A. cervicornis. Even limited unidirectional flow of genes may represent a significant mechanism for evolutionary change (Mallet 2005), thus by providing a conduit for gene flow; A. prolifera may provide a mechanism for the acquisition of novel genes for at least one (A. cervicornis) of the Caribbean Acropora. Interestingly, A. prolifera lacks a fossil record (Budd et al. 1994) suggesting that either hybridization between A. palmata and A. cervicornis is a relatively recent phenomenon or that environmental conditions have only recently favored survival of the hybrid A. prolifera. The current Caribbean-wide distribution of A. prolifera (Veron 2000, Wallace 1999) suggests that whatever triggers hybridization between its parental species has operated over a large geographic scale. by NORTHEASTERN UNIVERSITY on 01/11/07. For personal use only.

A Polyploid or Recombinational Hybrid Species? Annu. Rev. Ecol. Evol. Syst. 2006.37:489-517. Downloaded from arjournals.annualreviews.org Karyotype data for corals are few (Heyward 1985, Kenyon 1997); however, there is evidence of polyploidy in multiple species groups of Paciﬁc Acropora (Kenyon 1997). In several scleractinian genera (Acropora, Fungia, Gonipora, Lobophyllia, and Montipora), the diploid chromosomal number is 28 (Heyward 1985, Kenyon 1997). Of 22 Acropora species sampled, 3 apparent cases of polyploidy were identiﬁed—A. elseyi, A. ocellata, and A. valida. The last of these 3 species is a probable triploid, whereas the others are likely tetraploids (Kenyon 1997). Without additional molecular or morphological data, it is not clear if these 3 species are allopolyploid hybrid species or autopolyploids (i.e., polyploids produced by the same species). However,

www.annualreviews.org • Hybridization in Corals 501 ANRV292-ES37-18 ARI 17 October 2006 7:29

the number of polyploid species found in the single study and aforementioned evidence for hybridization implies the existence of polyploid hybrids among Indo-Pacific Acropora species and suggests that this phenomenon may not be uncommon among corals. Although no karyotype data are yet available for the three Caribbean Acropora, A. prolifera is unlikely to be polyploid because hybrids never had more than two alleles at single copy nuclear loci (S.V. Vollmer, unpublished data). Although known to be of hybrid origin, it is unclear whether A. prolifera is reproductively isolated from its parent species. In laboratory crosses, A. prolifera is both self-fertile and cross- fertile with A. cervicornis (backcrosses with A. palmata have not yet been completed; S.V. Vollmer, unpublished data). However, genetic evidence showing limited one- way gene flow from A. palmata to A. cervicornis implies that hybrids backcross with A. cervicornis at only low levels (Vollmer & Palumbi 2002). Both lines of evidence suggest that A. prolifera is not reproductively isolated from A. cervicornis. Moreover, the fact that all colonies of the hybrid A. prolifera so far tested appear to be F1 hybrids (Vollmer & Palumbi 2002) suggests that hybrids are not interbreeding at high frequency. Thus, the available genetic and reproductive evidence suggest that A. prolifera is a hybrid and not a hybrid species. Although A. prolifera may not strictly constitute a recombinational hybrid species, its ability to propagate clonally via asexual fragmentation allows these coral hybrids to persist locally, potentially over long periods of time (Vollmer & Palumbi 2002). The ability to persist through asexual propagation makes coral hybrids similar to clonal parthenogenic taxa (Miller & van Oppen 2003) that have been shown to be ecologically successful in a variety of animal groups (Bullini 1994, Vrijenhoek 1984). Indeed, A. prolifera, although generally rare, is most abundant in marginal and shallow water environments where neither of the parental species occurs (Cairns 1982; S.V. Vollmer, personal observation). Thus, these hybrids appear to be successful in an ecological niche that is distinct from that of the parent species. Sim- ilar habitat diversification is shown to be important in the evolution of plant hybrid species (Arnold 1997, Rieseberg et al. 2003). The recent detection of A. prolifera’s hybrid status raises the possibility that further molecular studies will reveal other coral morphospecies to be either hybrids or hybrid species that have diversi- fied to occupy new habitats (Vollmer & Palumbi 2002). Recent evidence that the by NORTHEASTERN UNIVERSITY on 01/11/07. For personal use only. European temperate soft coral Alcyonium hibernicum harbors the ITS variants of two congeners, A. coralloides and A. sp. M2, and consequently is a hybrid of the two species (McFadden & Hutchinson 2004) supports this view. Despite its exclu- Annu. Rev. Ecol. Evol. Syst. 2006.37:489-517. Downloaded from arjournals.annualreviews.org sively asexual mode of reproduction, A. hibernicum has a considerable distribution beyond the northern margins of the putative parent species’ distributions, highlighting the potential role that hybrid lineages may have for range expansion beyond the limits of parental species’ distributions. If climate change and other human-related pressures on Caribbean reefs continue at present levels or escalate, the survival of coral species may be increasingly dependent on their ability to colonize marginal habitats, highlighting the potentially increasing significance of hybrids and hybrid species.

502 Willis et al. ANRV292-ES37-18 ARI 17 October 2006 7:29

Para- and Polyphyly in Molecular Phylogenies of Indo-Pacific Mass-Spawning Species of Acropora In contrast to the low diversity of the Caribbean Acropora, there are more than 100 species of Indo-Pacific Acropora and up to 76 of these have been recorded from a single area (Wallace 1999). This complexity effectively precludes direct extrapolation of conclusions drawn on the basis of the much simpler, three-species Caribbean system to the Indo-Pacific Acropora. In particular, unlike the Caribbean fauna, there are no clear cases of species with morphologies intermediate between two putative parent species and, so far, molecular approaches have identified only two likely first- or early- generation hybrids (van Oppen et al. 2002, Wolstenholme 2004). The lack of obvious morphological hybrids may be partly explained by high levels of morphological variation within many coral species, which makes it difficult to locate species borders within morphological space, a prerequisite for distinguishing potential hybrids from intraspecific variation. Recently, however, morphs that have affinities to two species have been identified within the A. humilis group (Wolstenholme 2004, Wolstenholme et al. 2003). Evidence from molecular and reproductive studies suggests that species boundaries may be at various stages of formation among a group of five species and seven intermediate morphs and that some of this morphological variation may have arisen through hybridization (e.g., morphs intermediate between A. digitifera and A. gemmifera; Wolstenholme 2004). In contrast to the Caribbean, where opportunities for hybridization are limited to two species, in the Indo-Pacific, introgression between any pair of species can theoretically occur through backcrossing with a range of other species (and intermediate morphs). The high number of potentially interbreeding species, combined with intraspecific morphological variation, poses difficulties for unraveling molecular phylogenies within the Indo-Pacific Acropora. Ribosomal DNA ITS phylogenies for Indo-Pacific Acropora spp. are consistent with interspecific hybridization and introgression in that they show sharing of highly divergent sequence types between a wide range of species (Marquez et al. 2003, Odorico & Miller 1997, van Oppen et al. 2002). A divergence of more than 20% has been observed between ITS1-5.8S-ITS2 sequences within species and even within individual corals, while at the same time some sequence variants show high similarity between species. In the A. aspera species group, one species, A. aspera, differs at least

by NORTHEASTERN UNIVERSITY on 01/11/07. For personal use only. in some years in the timing of gamete release and harbors mainly a distinct and unique ITS type (van Oppen et al. 2002). The other four species that have been examined in this group spawn simultaneously every year and also share ITS types, supporting

Annu. Rev. Ecol. Evol. Syst. 2006.37:489-517. Downloaded from arjournals.annualreviews.org the view that synchronized spawning has led to hybridization and the sharing of ITS types. Furthermore, there is some evidence that some A. pulchra individuals may be recent generation hybrids (van Oppen et al. 2002). As ITS homogenization may proceed extremely slowly when divergent sequence types are combined in a single genome (Modrich & Lahue 1996), rDNA data cannot readily distinguish between ancient or recent hybridization events. Incomplete lineage sorting may further mimic hybridization signatures (Vollmer & Palumbi 2004). Nevertheless, phylogenies based on single-copy nuclear DNA (scnDNA) and mtDNA are broadly consistent with the rDNA analyses, also showing extended para- and polyphyly of a wide range of species

www.annualreviews.org • Hybridization in Corals 503 ANRV292-ES37-18 ARI 17 October 2006 7:29

and resolving A. aspera from other members of the A. aspera species group (Hatta et al. 1999, van Oppen et al. 2001). Within the A. humilis group, sequence data from the partial nuclear ribosomal large subunit DNA regions and the mitochondrial control region suggest that sequence types are shared between some species through occasional introgression without dis- rupting morphological boundaries (Wolstenholme 2004, Wolstenholme et al. 2003). Morphs intermediate between true species are possibly of hybrid origin based on fertilization tests, their morphological affinities, and their phylogenetic position in molecular phylogenies (Wolstenholme 2004, Wolstenholme et al. 2003). Paraphyly has also been observed for several other Acropora species at scnDNA and mtDNA markers (Hatta et al. 1999, Marquez´ et al. 2002b, van Oppen et al. 2001). In addition, some aspects of mtDNA and scnDNA phylogenies are inconsistent. Such patterns could be explained by interspecific hybridization and introgression, incomplete lineage sorting (i.e., shared ancestral polymorphism), or a combination of both. Two lines of evidence, however, lead us to believe that hybridization is at least partly responsible. First, as enumerated above, many mass-spawning species of Indo-Pacific Acropora are capable of successful cross-fertilization (Fukami et al. 2003, Hatta et al. 1999, van Oppen et al. 2002, Willis et al. 1997), and second, the genetic distinctiveness of a nominal species appears to be directly correlated with the extent of temporal or other reproductive barriers (Fukami et al. 2003, van Oppen et al. 2001, 2002, Wolstenholme 2004), with species having the most effective temporal isolating mechanisms containing distinct rDNA ITS1 variants, scnDNA alleles, and mtDNA haplotypes. For example, A. donei and A. yongei in Japan, which spawn 1–3 hours before other sympatric species, constitute a genetically distinct cluster based on Mini-C intron 2 sequences (Fukami et al. 2003, Hatta et al. 1999). A. digitifera on the GBR, which spawns three months after the main mass-spawning event, constitutes a separate clade from the other four species in the A. humilis group based on analyses of the mtDNA intergenic region (Wolstenholme 2004). Spawning of A. yongei in the central GBR was correctly inferred to coincide with that of A. tenuis, which similarly spawns 2–3 hours before congeners (Babcock et al. 1986) based on its phylogenetic position in the PaxC intron and control region trees (M.J.H. van Oppen, B.L. Willis, D.J. Miller, unpublished data). Moreover, ITS diversity correlates with the permeability of isolating mechanisms; ITS1 variability within and between most Indo-Pacific Acro- by NORTHEASTERN UNIVERSITY on 01/11/07. For personal use only. pora species is much higher than that observed among the three-species Caribbean A. cervicornis group. Thus, ITS1 p-distances range from 0% to ∼13% in the A. cervicornis group, but are up to four times greater in the Indo-Pacific A. aspera group. A. Annu. Rev. Ecol. Evol. Syst. 2006.37:489-517. Downloaded from arjournals.annualreviews.org tenuis has even lower ITS1 diversity (0–5.3%; M.J.H. van Oppen, unpublished data) and is genetically distinct based on scnDNA and mtDNA, which is consistent with its earlier spawning time and low interspecies fertilization success in vitro (Babcock et al. 1986, Willis et al. 1997). Similarly, A. latistella typically (but not always) spawns two weeks out of phase with most congeners (Babcock et al. 1986, Willis et al. 1985), and this species as well as A. aspera (which may spawn earlier than the mass spawning) both form distinct clusters in the molecular analyses. Thus transient reproductive barriers imposed by year-to-year variation in the lunar night of spawning may also contribute to the relative genetic distinction of Acropora species. In summary, a clear

504 Willis et al. ANRV292-ES37-18 ARI 17 October 2006 7:29

pattern linking lack of temporal barriers to breeding with nonmonophyly in molecular phylogenies helps to build the case that hybridization has contributed to the evolution of the mass-spawning, Indo-Paciﬁc Acropora.

Population Genetic Approaches Show that Some Nonmonophyletic Species are Genetically Distinct Molecular phylogenies are extremely powerful tools for identifying likely cases of natural hybridization, but additional approaches are required to corroborate the occurrence and extent of introgressive hybridization. Although allele sharing in phylogenetic trees of the Indo-Pacific Acropora is strongly indicative of natural hybridization and introgression within the genus, such evidence alone cannot distinguish occasional hybridization between true species from the existence of a single, morphologically polymorphic species or from cases of incomplete lineage sorting. Because differences in allele frequencies occur more rapidly than mutational changes underlying phylogenetic analyses, population genetic approaches provide further insights into whether closely related cross-fertile species share the same gene pool or are connected through limited hybridization and introgression. Population genetic studies of two corals, A. cytherea and A. hyacinthus, which have highest interspecific fertility in experimental crosses (Willis et al. 1997) and share alleles extensively in phylogenetic analyses (Marquez´ et al. 2002b), demonstrate that these two taxa do not represent a single morphologically plastic species (Marquez´ et al. 2002a). Although very low, levels of genetic differentiation between these two species in sympatry were significant at eight polymorphic allozyme loci. Higher levels of gene flow between conspecific, widely allopatric (eastern versus western Australian) populations of these species in comparison to interspecific gene flow between them in sympatry support the conclusion that A. cytherea and A. hyacinthus constitute distinct entities (Marquez´ et al. 2002a). The question of whether incomplete lineage sorting is responsible for the small genetic divergences is more difficult to address. Based on the absence of a fossil record for either A. cytherea or A. hyacinthus prior to the Pleistocene (Wallace 1999), both species are assumed to be of relatively recent origin. The lack of fixed allelic differences and low allele frequency differences between these two species are both consistent with natural hybridization and introgression

by NORTHEASTERN UNIVERSITY on 01/11/07. For personal use only. (i.e., interspeciﬁc gene ﬂow following secondary contact) occurring infrequently, but could also be explained by them being incipient species with incomplete reproductive barriers and retention of ancestral polymorphisms (Marquez´ et al. 2002b).

Annu. Rev. Ecol. Evol. Syst. 2006.37:489-517. Downloaded from arjournals.annualreviews.org Low, but distinct genetic differentiation in population genetic studies of the A. nasuta group (MacKenzie 2005) provides a second example of infrequent hybridization events potentially contributing to the evolution of species in the Indo-Paciﬁc Acro- pora. Three species in this group, A. nasuta, A. valida, and A. secale, showed highly signiﬁcant indices of pairwise genetic differentiation, both in sympatry and in al-

lopatry (Fst values based on two nuclear introns and one microsatellite locus ranged from 0.08–0.09, 0.24–0.37, and 0.38–0.42 between A. secale and A. valida, A. nasuta and A. valida, and A. nasuta and A. secale, respectively) (MacKenzie 2005). Of the three species studied, two (A. nasuta and A. valida) were included in the PaxC intron phylogeny (van Oppen et al. 2001) and showed para- and polyphyly, respectively.

www.annualreviews.org • Hybridization in Corals 505 ANRV292-ES37-18 ARI 17 October 2006 7:29

These examples illustrate that nominal Acropora species constitute genetically distinct entities, some of which are likely to exchange genes with congeneric species (as recognized within the current taxonomic framework; Wallace 1999) at low frequencies through introgressive hybridization. Thus, although no-choice cross-fertilization trials demonstrate mating compatibilities (Hatta et al. 1999, Miller & Babcock 1997, Szmant et al. 1997, van Oppen et al. 2002, Willis et al. 1997), sperm-choice trials are a much better predictor of the extent of hybridization occurring under natural conditions, where eggs are exposed to complex mixtures of con- and heterospeciﬁc sperm that compete to fertilize eggs. Nevertheless, even rare hybridization events on ecological timescales are likely to be signiﬁcant on evolutionary timescales.

Hybridization in Other Mass-Spawning and Brooding Corals There have been few genetic studies of mating systems in other mass-spawning coral genera on Indo-Pacific reefs, but the one study that has explored the potential for hybridization in a nonacroporid coral found evidence of extensive interspecific breeding. The genus Platygyra (Faviidae) on the GBR comprises seven recognized morphospecies, and estimates of pairwise genetic differences among species based on allozymes are low (Miller & Benzie 1997). Although corroborative evidence from additional molecular markers is required, low allozyme divergence (Nei’s D range from 0.032 to 0.057) combined with the fact that the species show overlap in spawning times and are reproductively compatible in experimental breeding trials (Miller & Babcock 1997) suggest that introgressive hybridization has also played a role in the evolution of this genus. In contrast to the Indo-Pacific Acropora, however, natural hybridization may be operating to homogenize morphospecies within the Platygyra. Alternatively, recent speciation has occurred with or without ongoing hybridization (Miller & Benzie 1997). Molecular studies of Caribbean species in the genus Madracis suggest that hybridization may also occur among brooding corals. Mixed paraphyly and site-specific polymorphisms, including additivity at nine sites, were interpreted as provisional evidence of hybridization among M. decactis, M. pharensis, and M. formosa (Diekmann et al. 2001). Interestingly, a recent allozyme study in the Indo-Pacific of two brooding scleractinian corals, Pocillopora damicornis and Stylophora pistillata, at the edge of their

by NORTHEASTERN UNIVERSITY on 01/11/07. For personal use only. ranges on high latitude reefs of Lord Howe Island (beyond the southern extremity of the GBR), revealed a small proportion of apparently introgressed hybrids (Miller & Ayre 2004). Such hybridization among brooding coral species may yet prove to be

Annu. Rev. Ecol. Evol. Syst. 2006.37:489-517. Downloaded from arjournals.annualreviews.org as widespread as for mass-spawning species. Alternatively, such events may be more prevalent in areas such as Lord Howe Island, where conspeciﬁc sperm may occur at unusually low densities.

COMPARATIVE BIOGEOGRAPHIC PATTERNS AND THE GEOGRAPHIC SCALE OF HYBRIDIZATION Molecular studies reviewed in the above sections demonstrate that hybridization events have occurred throughout the evolutionary history of corals on both Caribbean and Indo-Paciﬁc reefs; however, the frequency and outcomes of hybridization appear

506 Willis et al. ANRV292-ES37-18 ARI 17 October 2006 7:29

to vary between the two regions and between genera. In the Indo-Pacific Acropora, allele and haplotype sharing indicate that hybridization occurs between many pairs of species on evolutionary timescales, but these hybridization events appear to be rare on ecological timescales. Because a few hybrids can provide a means for transferring alleles between species, even rare events on ecological timescales may represent a significant mechanism for evolutionary change (Mallet 2005). Such events may have contributed significantly to adaptive radiation of the Indo-Pacific Acropora, particularly as they colonized new shallow-water habitats on continental shelves following frequent sea-level transgressions during the Pleistocene (Veron 1995). In contrast, hybridization between two of the three Caribbean Acropora species appears to be common, resulting in a hybrid that is recognized as the separate morphospecies, A. prolifera, although the outcome of hybridization appears to be limited to the produc- tion of F1 hybrids. One factor that may contribute significantly to apparent differences in the frequency of hybridization is the comparative species richness of these two biogeographic regions. In the Caribbean, there are only three species within the genus Acropora and the coral fauna overall is quite small; around 30 scleractinian genera are present (Veron 2000), but each characteristically contains few species. This con- trasts markedly with the Indo-Pacific fauna, where there are more than 100 species of Acropora and more than 80 genera overall (Veron 2000). When many congeneric species spawn simultaneously in sympatry, there may be stronger selection for effi- cient gamete recognition, whereas in locations where gametes of fewer species have an opportunity to interact, selection may be less stringent. However, as discussed above, even on Indo-Pacific reefs, isolating mechanisms are not absolute; many species of Acropora retain options for opportunistic interspecific mating. Limited evidence suggests that there may also be variation in the frequency of hybridization within biogeographic regions. Hybridization appears to be more frequent at the edges of some coral species’ distributional ranges, for example, at the northern periphery of the distribution of the Caribbean genus Montastraea (Fukami et al. 2004). There is also limited evidence of intergeneric hybridization between Pocillopora and Stylophora at the southern extremity of their ranges on the GBR (Miller & Ayre 2004). In such peripheral locations, densities of mates are likely to be low, few species are present, and hybrids may be able to exploit nonparental niches, as shown for plant species (Lexer et al. 2003, Rieseberg et al. 2003). by NORTHEASTERN UNIVERSITY on 01/11/07. For personal use only. The broad geographic scales over which corals interact through introgressive hybridization are unparalleled in terrestrial animal groups. The distribution of the coral hybrid Acropora prolifera throughout the Caribbean (Veron 2000, Wallace 1999) Annu. Rev. Ecol. Evol. Syst. 2006.37:489-517. Downloaded from arjournals.annualreviews.org suggests that hybridization occurs over the entire biogeographic ranges of the two parental species, A. palmata and A. cervicornis. Similarly, species of Indo-Pacific Acro- pora are sympatric over hundreds to thousands of kilometers and have consistent patterns of low interspecific gene flow on both the eastern and western coasts of Australia (Marquez´ et al. 2002a). Other pairs of Acropora species have a reticulate evolutionary history on reefs in Japan (Hatta et al. 1999). Interbreeding on this scale differs radically from the norm for terrestrial animals, where hybrid zones are usu- ally narrow and maintained by a balance between dispersal and selection (reviewed in Arnold 1997, Barton & Hewitt 1985). Although concepts of hybrid zones have

www.annualreviews.org • Hybridization in Corals 507 ANRV292-ES37-18 ARI 17 October 2006 7:29

expanded to encompass mosaic hybrid zones and attention is now focusing on geographical distributions of genealogical lineages (reviewed in Hewitt 2001), in general, animal hybrids have been assumed to be limited to contact zones where neither parent species is particularly ﬁt. The closest parallels to potential geographic scales of hybridization among corals are found in plants. For example, oak species in the genus Quercus in the eastern United States (Whittemore & Schaal 1991) and Eucalyptus species in Australia (Potts & Reid 1988) are accepted as syngameons of interbreeding species across broad distributional ranges. Recent evidence that hybridizing species of coral reef ﬁsh are also broadly sympatric but ecologically partitioned by reef habitat (van Herwerden et al. 2006) suggests that broad scales of interbreeding may not be uncommon among broadcast-spawning animals.

SPECIES CONCEPTS FOR HYBRIDIZING CORALS Coral species are traditionally recognized on the basis of morphological characters, particularly skeletal structures, but evolutionary relationships implied by morphology are not always reflected in molecular phylogenies (Fukami et al. 2004, van Oppen et al. 2001, Wolstenholme et al. 2003). Different rates of evolution for morphological and molecular characters (Wolstenholme et al. 2003), phenotypic plasticity, and con- vergent evolution may partly explain these mismatches, but reticulate evolutionary relationships detected among mass-spawning corals of Indo-Pacific Acropora (Hatta et al. 1999; van Oppen et al. 2001, 2002) are also likely to have contributed, particularly in cases of closely related species that interbreed. Regardless of such mismatches, it is clear that discrete coral species with distinct ecological characteristics do exist and define stable and cohesive entities despite occasional interspecific gene flow (Marquez´ et al. 2002a, Miller & Babcock 1997, Wolstenholme et al. 2003, as described above for A. millepora and A. pulchra). Examples of interspecific gene exchange occurring between corals that are nonetheless morphologically and ecologically distinct suggest that, just as in plants, cohesive mechanisms override occasional gene flow to maintain recognizable species boundaries. Spatial heterogeneity on reefs, particularly as caused by steep environmental gradients (e.g., light, hydrodynamic and exposure gradients with depth), leads to dramatically different habitats over scales measured in tens of meters and provides by NORTHEASTERN UNIVERSITY on 01/11/07. For personal use only. a mechanism by which disruptive selection could maintain morphologically discrete taxa with distinct physiological requirements over distances small enough to allow mixing of gametes. Thus, even when adults of mass-spawning species differ in their Annu. Rev. Ecol. Evol. Syst. 2006.37:489-517. Downloaded from arjournals.annualreviews.org benthic habitats along depth-defined environmental gradients, gametes from these species are able to interact and fertilize at the sea surface, continually renewing the potential for interspecific gene flow. Disruptive selection combined with disjunction between the benthic habitat of adults and the sea surface location of fertilization may partly explain how well-defined morphological and ecological coral species persist while maintaining opportunities for some interspecific gene flow. Because many coral species share genetic variation with closely related species, phylogenetic species concepts based on species monophyly (e.g., Baum & Shaw 1995, Cracraft 1989) should not be applied to corals, especially because genes can have

508 Willis et al. ANRV292-ES37-18 ARI 17 October 2006 7:29

different histories of separation and gene flow (e.g., within the Caribbean Acropora; Vollmer & Palumbi 2002). A more appropriate species concept for corals is one that embodies the spirit of the biological species concept (sensu Dobzansky 1937, Mayr 1963) in which species are viewed as entities that are effectively reproductively isolated from other such groups (i.e., the cohesion species concept of Templeton 1989), but acknowledges that complete reproductive isolation among species does not reflect biological reality. The approach of plant biologists, who have always tolerated a degree of hybridization among well-defined species (sensu Grant 1981, Rieseberg 1997), is particularly relevant to coral species. Similarly, the botanically derived concept of a syngameon (Grant 1981), which links species that are capable of interbreeding and exchanging genes in a reproductive community, usefully describes many groups of closely related coral species. The possession of a mating system that tolerates and potentially takes advantage of a degree of interspecific gene exchange adds a new dimension to the similarities between the life histories of corals and plants. In view of this, delimiting good coral species requires multiple lines of evidence (e.g., as for the Caribbean Montastraea and Indo-Pacific Acropora) to interpret how sometimes contradictory morphological, ecological, reproductive, and genetic characters best characterize cohesive coral species (Wallace & Willis 1994). Although the extent to which the consequences of gene flow are potentially adaptive versus negative requires further exploration, this propensity of corals adds to current understanding of the role of hybridization in animal speciation.

SIGNIFICANCE OF HYBRIDIZATION FOR TROPICAL CORAL SPECIES AND THE RESILIENCE OF CORAL REEFS Tropical reef corals are facing a suite of challenges, the cumulative impact of which may jeopardize the persistence of some (e.g., Caribbean) coral reef ecosystems as they have been known in recent geological times. The threat of ocean warming is potentially catastrophic to animals that live within 1–2◦C of their upper thermal limits (Hoegh-Guldberg 1999, Hughes et al. 2003). Moreover, exploitation of coral reef communities has disturbed their trophic balance, in some cases causing community phase shifts from coral- to algal-dominated reefs (Hughes et al. 2003). The unprecedented rise of coral disease on Caribbean reefs has decimated populations of by NORTHEASTERN UNIVERSITY on 01/11/07. For personal use only. the dominant framework-building corals in the genera Acropora and Montastraea, as well as major gorgonian species (Harvell et al. 2002). Torespond to these challenges, coral species need a variety of mechanisms for rapid evolutionary change. Annu. Rev. Ecol. Evol. Syst. 2006.37:489-517. Downloaded from arjournals.annualreviews.org Hybridization events are an important source of raw material for rapid evolutionary change in a variety of plant and animal groups and also have the potential to facilitate adaptive radiation when new adaptive zones are invaded (Seehausen 2004). A study of Darwin’s finches in the Galapagos Islands highlights the ability of hybrids to provide a rapid response to environmental change, in this case through novel beak morphologies when climatological changes abruptly reduced the seed types available (Grant & Grant 1996). Moreover, backcrossing facilitated the persistence of the parental species’ genetic diversity when finches failed to breed successfully in purebred matings. Hybridization may also significantly increase resilience to novel disease

www.annualreviews.org • Hybridization in Corals 509 ANRV292-ES37-18 ARI 17 October 2006 7:29

challenges. For example, hybrids of frog species arising through allopolyploid speciation had increased disease resistance to two different parasites that each infected one of the parent species ( Jackson & Tinsley 2003). Similarly, hybrid parakeets had significantly higher measures of immune function than inbred parental species, through enrichment of depauperate gene pools (Tompkins et al. 2006). Such evidence that hybridization can mediate the response of species to rapid climatological change and the development of disease resistance highlights the potential significance hybridization may hold for the persistence of coral species, which face similar environmental challenges. In conclusion, hybridization has contributed in diverse and significant ways to the evolution of coral species in the ecologically prominent Indo-Pacific Acropora,as well as to the major Caribbean framework-building genera Montastraea and Acropora. We assert that the evolutionary potential of hybridization is important to conserve, thus hybrids like A. prolifera represent important reservoirs of novel genetic diversity that may facilitate adaptive radiation under changed environmental circumstances. In addition, their ability to pass genes between species represents a potentially significant mechanism for rapid evolutionary change of parental species. As advocated by Ennos et al. (2005) in a recent review of conservation principles for taxonomically complex groups, efforts should be directed toward the conservation of evolutionary processes that generate such biodiversity rather than the current focus on the conservation of species, which are ambiguous entities in taxonomically complex groups. Recommendations against the need to protect hybridizing species because they fall outside the framework of conventional species definitions or are interpreted as jeop- ardizing the persistence of pure native species (e.g., Allendorf et al. 2001) ignore the dynamic nature of species and the potentially integral role that hybridization may play in the founding of lineages. Given the implications of ocean warming for increasing frequency and severity of mass bleaching events and emerging coral diseases, it is important to realize that hybridization has contributed to the evolutionary diversification of corals and that it has the potential to contribute significantly to the resilience of coral species and the coral reef ecosystem into the future. Conservation strategies that protect the evolutionary processes that have given rise to modern reef corals and are likely to contribute to their future resilience should be paramount. by NORTHEASTERN UNIVERSITY on 01/11/07. For personal use only.

SUMMARY POINTS

Annu. Rev. Ecol. Evol. Syst. 2006.37:489-517. Downloaded from arjournals.annualreviews.org 1. Hybridization has contributed to the evolution of many ecologically dominant and structurally important corals in diverse and significant ways. 2. Interspecific gamete compatibility in no-choice crosses and developmental competence of hybrids suggest there may be few absolute pre- or postzygotic barriers to interspecific breeding among many mass-spawning species of Indo-Pacific Acropora. However, in sperm choice experiments, conspecific matings take precedence, providing a mechanism for maintaining morphospecies boundaries when prezygotic barriers are semi-permeable.

510 Willis et al. ANRV292-ES37-18 ARI 17 October 2006 7:29

3. Although decoupling of adult (benthic) and natal (sea surface) habitats in mass-spawning corals continually renews opportunities for hybridization, spatial heterogeneity and steep, depth-related environmental gradients on reefs provide a mechanism by which disruptive selection could also contribute to the maintainance of morphologically and ecologically discrete taxa despite occasional gene flow. 4. Juvenile coral hybrids are not less fit than purebreds, as indicated by comparative survival and growth in a large-scale grow-out program. Greater growth and survival of juvenile hybrids in environmentally variable and extreme habitats suggest a role for hybrids in adaptation to new environments. 5. In the Indo-Pacific, molecular phylogenies of mass-spawning species in the genus Acropora are consistent with reticulate evolutionary pathways. Al- though population genetic studies indicate hybridization events are rare on ecological timescales, on evolutionary timescales, they are likely to have facilitated adaptive radiations leading to their current high diversity. 6. In the comparatively depauperate Caribbean, molecular studies reveal that one of only three exant Atlantic Acropora species is a hybrid, providing a conduit for one-way gene flow from A. palmata to A. cervicornis and a mechanism for the introduction of novel genetic material. 7. The capacity of the hybrid A. prolifera to colonize marginal habitats distinct from its parent species and evidence of hybridization at geographical boundaries of the Caribbean Montastraea support an evolutionary role for hybridization in range expansion and adaptation to changing environments. 8. The distribution of the hybrid A. prolifera throughout the entire Caribbean ranges of its parent species highlights the broad geographic scales of hybridization possible in marine environments when the location of gamete interactions is spatially segregated from adult habitats. Such scales differ from the narrow hybrid zones typical of most animals and provide new insights into the process of hybridization in animal species evolution.

by NORTHEASTERN UNIVERSITY on 01/11/07. For personal use only. 9. In combination, outcomes of hybridization are likely to be signiﬁcant for the future resilience of reef corals, for example, by providing options for rapid response to changing environments and climatologies as well as increasing

Annu. Rev. Ecol. Evol. Syst. 2006.37:489-517. Downloaded from arjournals.annualreviews.org resilience to novel disease challenges. Hybridization warrants consideration when developing conservation strategies for reef corals.

ACKNOWLEDGMENTS We thank D. Thomson, A. Beer, and B. Miller for ﬁeld assistance; F. van der Leest for data management in the hybrid and purebred outplanting study; R. Standish for assistance with allozyme studies; J. Veron, H. Ruiz, and F. Seneca for use of images in Figures 1a and 1b, respectively; D. Harvell and L. van Herwerden for useful

www.annualreviews.org • Hybridization in Corals 511 ANRV292-ES37-18 ARI 17 October 2006 7:29

comments on the manuscript; and R. Willis for editorial assistance. This work was supported by funds from the Australian Research Council and James Cook University.

LITERATURE CITED Acropora Biological Review Team.2005. Atlantic Acropora status review document. Re- port to National Marine Fisheries Service, Southeast Regional Office. March 3, 2005. 152 pp. + Appendix Adey WH, Gladfelter W, Odgen J, Dill R. 1977. Field guidebook to the reefs and reef communities of St. Croix, Virgin Islands. Third Int. Symp. Fla: Coral Reefs, The Atlantic Reef Comm., Univ. Miami Allendorf FW, Leary RF, Spruell P, Wenburg JK. 2001. The problems with hybrids: setting conservation guidelines. Trends Ecol. Evol. 16:613–22 Anderson E. 1949. Introgressive Hybridization. New York: Wiley Anderson E, Stebbins GL. 1954. Hybridization as an evolutionary stimulus. Evolution 8:378–88 Arnold ML. 1997. Natural Hybridization and Evolution. New York:Oxford Univ. Press. 215 pp. Arnold ML, Hodges SA. 1995. Are natural hybrids fit or unfit relative to their parents? Trends Ecol. Evol. 10:67–71 Babcock RC, Bull GD, Harrison PL, Heyward AJ, Oliver JK, et al. 1986. Mass spawning of 105 scleractinian coral species on the Great Barrier Reef. Marine Biol. 90:379–94 Barton NH, Hewitt GM. 1985. Analysis of hybrid zones. Annu. Rev. Ecol. Syst. 16:113– 48 Baird AH, Marshall PA, Wolstenholme J. 2002. Mass spawning of Acropora in the Coral Sea. Proc. Ninth Int. Coral Reef Symp. Bali. 1:385–89 Baum DA, Shaw KL. 1995. Genealogical perspectives on the species problem. In Experimental and Molecular Approaches to Plant Biosystematics, ed. PC Hoch, AG Stevenson, pp. 289–303. St. Louis: Missouri Bot. Gard. Budd AF, Johnson KG. 1999. Origination preceding extinction during late Cenozoic turnover of Caribbean reefs. Paleobiology 25:188–200 Budd AF, Pandolfi JM. 2004. Overlapping species boundaries and hybridization by NORTHEASTERN UNIVERSITY on 01/11/07. For personal use only. within the Montastraea “annularis” reef coral complex in the Pleistocene of the Bahama Islands. Paleobiology 30:396–425 Budd AF, Stemann TA, Johnson KG. 1994. Stratigraphic distributions of genera and Annu. Rev. Ecol. Evol. Syst. 2006.37:489-517. Downloaded from arjournals.annualreviews.org species of Neogene to recent Caribbean reef corals. J. Paleontol. 68:951–77 Buerkle CA, Morris RJ, Asmussen MA, Rieseberg LH. 2000. The likelihood of homoploid hybrid speciation. Heredity 84:441–51 Bullini L. 1994. Origin and evolution of animal hybrid species. Trends Ecol. Evol. 9:422–26 Cairns SD. 1982. Stony corals (Cnidaria; Hydrozoa, Scleractinia) of Carrie Bow Cay, Belize. See Rutzler¨ & Macintyre 1982, pp. 271–302 Carroll A, Harrison P, Adjeroud M. 2006. Sexual reproduction of Acropora reef corals at Moorea, French Polynesia. Coral Reefs 25:93–97

512 Willis et al. ANRV292-ES37-18 ARI 17 October 2006 7:29

Coyne JA, Orr HA. 2004. Speciation. Sunderland: Sinauer. 545 pp. Cracraft J. 1989. Speciation and its ontology: the empirical consequences of alternative species concepts for understanding patterns and processes of differentiation. In Speciation and its Consequences, ed. D Otte, JA Endler, pp. 28–59. New York: Sinauer de Graaf M, Geertjes GJ, Videler JJ. 1999. Observations on spawning of scleractinian corals and other invertebrates on the reefs of Bonaire (Netherlands Antilles, Caribbean). Bull. Mar. Sci. 64:189–94 Diekmann OE, Bak RPM, Stam WT, Olsen JL. 2001. Molecular genetic evidence for probable reticulate speciation in the coral genus Madracis from a Caribbean fringing reef slope. Mar. Biol. 139:221–33 Dobzhansky T. 1937. Genetics and the Origin of Species. New York: Columbia Univ. Press. 364 pp. Dowling TE, Secor CL. 1997. The role of hybridization and introgression in the diversification of animals. Annu. Rev. Ecol. Syst. 28:593–619 Ennos RA, French GC, Hollingsworth PM. 2005. Conserving taxonomic complexity. Trends Ecol. Evol. 20:164–68 Freeland JR, Boag PT. 1999. The mitochondrial and nuclear genetic homogeneity of the phenotypically diverse Darwin’s ground finches. Evolution 53:1553–63 Fukami H, Budd AF, Levitan DR, Jara J, Kersanach R, Knowlton N. 2004. Geo- graphic differences in species boundaries among members of the Montastraea annularis complex based on molecular and morphological markers. Evolution 58:324–37 Fukami H, Omori M, Shimoike K, Hayashibara T, Hatta M. 2003. Ecological and genetic aspects of reproductive isolation by different spawning times in Acropora corals. Mar. Biol. 142:679–84 Gardner JPA. 1997. Hybridization in the sea. Adv. Mar. Biol. 31:1–78 Gittings SR, Boland GS, Deslarzes KJ, Combs CL, Holland BS, Brights TJ. 1992. Mass spawning and reproductive variability of reef corals at the East Flower Garden Bank, north-west Gulf of Mexico. Bull. Mar. Sci. 51:420–28 Goreau TF. 1959. The ecology of Jamaican coral reefs I: Species composition and zonation. Ecology 40:67–90 Grant V. 1981. Plant Speciation. New York: Columbia Univ. Press. 563 pp. 2nd ed. Grant PR, Grant R. 1996. High survival of Darwin’s finch hybrids: effects of beak by NORTHEASTERN UNIVERSITY on 01/11/07. For personal use only. morphology and diet. Ecology 77:500–9 Gregory JW. 1895. Contributions to the palaeontology and physical geology of the

Annu. Rev. Ecol. Evol. Syst. 2006.37:489-517. Downloaded from arjournals.annualreviews.org West Indies. Q. J. Geol. Soc. London 51:255–312 Guest JR, Chou LM, Baird AM, Goh BPL. 2002. Multispeciﬁc, synchronous coral spawning in Singapore. Coral Reefs 21:422–23 Harrison PL, Babcock RC, Bull GD, Oliver JK, Wallace CC, Willis BL. 1984. Mass spawning in tropical reef corals. Science 223:1186–89 Harrison PL, Wallace CC. 1990. Reproduction, dispersal and recruitment of scleractinian corals. In Ecosystems of the World, 25: Coral Reefs, ed. Z Dubinsky, pp 133–207. Amsterdam: Elsevier Harvell CD, Mitchell CE, Ward JR, Altizer S, Dobson AP, et al. 2002. Climate warming and disease risks for terrestrial and marine biota. Science 296:2158–62

www.annualreviews.org • Hybridization in Corals 513 ANRV292-ES37-18 ARI 17 October 2006 7:29

Hatta M, Fukami H, Wang W, Mori M, Shimoike K, et al. 1999. Reproductive and genetic evidence for a reticulate evolutionary history of mass-spawning corals. Mol. Biol. Evol. 16:1607–13 Hayashibara T, Shimoike K, Kimura T, Hosaka S, Heyward A, et al. 1993. Patterns of coral spawning at Akajima Island, Okinawa, Japan. Mar.Ecol. Prog. Ser. 10:253–62 Hewitt GM. 2001. Speciation, hybrid zones and phylogeography—or seeing genes in space and time. Mol. Ecol. 10:537–49 Heyward AJ. 1985. Comparative coral karyology. Proc. Fifth Int. Coral Reef Cong. Tahiti 6:47–51 Hoegh-Guldberg O. 1999. Coral bleaching, climate change and the future of the world’s coral reefs. Mar. Freshw. Res. 50:839–66 Hughes TP, Baird AH, Bellwood DR, Card M, Connolly SR, et al. 2003. Climate change, human impacts and the resilience of coral reefs. Science 301:929–33 Jackson JA, Tinsley RC. 2003. Parasite infectivity to hybridizing host species: a link between hybrid resistance and allopolyploid speciation? Int. J. Parasitol. 33:137– 44 Jiggins CD, Mallet J. 2000. Bimodal hybrid zones and speciation. Trends Ecol. Evol. 15:250–55 Kearney M. 2005. Hybridization, glaciation and geographical parthenogenesis. Trends Ecol. Evol. 20:495–502 Kenyon JC. 1997. Models of reticulate evolution in the coral genus Acropora based on chromosome numbers: parallels with plants. Evolution 51:756–67 Knowlton N, Mate JL, Guzman HM, Rowan R, Jara J. 1997. Direct evidence for reproductive isolation among the three species of the Montastraea annularis complex in Central America (Panama and Honduras). Mar. Biol. 127:705–11 Lexer C, Welch ME, Raymond O, Rieseberg LH. 2003. The origin of ecological divergence in Helianthus paradoxus (Asteraceae): selection on transgressive characters in a novel hybrid habitat. Evolution 57(9):1989–2000 Levitan DR, Fukami H, Jara J, Kline D, McGovern TM, et al. 2004. Mechanisms of reproductive isolation among sympatric broadcast-spawning corals of the Mon- tastraea annularis species complex. Evolution 58:308–23 MacKenzie JB. 2005. Population-species architecture of corals in the Acropora nasuta group. PhD thesis. James Cook Univ., Townsville, Aust. Mallet J. 2005. Hybridization as an invasion of the genome. Trends Ecol. Evol. 20:229– by NORTHEASTERN UNIVERSITY on 01/11/07. For personal use only. 37 Marquez´ LM, Miller DJ, MacKenzie JB, van Oppen MJH. 2003. Pseudogenes con-

Annu. Rev. Ecol. Evol. Syst. 2006.37:489-517. Downloaded from arjournals.annualreviews.org tribute to the extreme diversity of nuclear ribosomal DNA in the hard coral Acropora. Mol. Biol. Evol. 20:1077–86 Marquez´ LM, van Oppen MJH, Willis BL, Miller DJ. 2002a. Sympatric populations of the highly cross-fertile coral species Acropora hyacinthus and A. cytherea are genetically distinct. Proc. R. Soc. London Ser. B 269:1289–94 Marquez´ LM, van Oppen MJH, Willis BL, Reyes A, Miller DJ. 2002b. The highly cross-fertile coral species, Acropora hyacinthus and A. cytherea, constitute statisti- cally distinguishable lineages. Mol. Ecol. 11:1339–49 Mateos M, Vrijenhoek RC. 2002. Ancient versus reticulate origin of a hemiclonal lineage. Evolution 56:985–92

514 Willis et al. ANRV292-ES37-18 ARI 17 October 2006 7:29

Mayr E. 1963. Animal Species and Evolution. Cambridge: Harvard Univ. Press. 797 pp. McCarthy EM, Asmussen MA, Anderson WW. 1995. A theoretical assessment of recombinational speciation. Heredity 74:502–09 McFadden CS, Hutchinson MB. 2004. Molecular evidence for the hybrid origin of species in the soft coral genus Alcyonium (Cnidaria: Anthozoa: Octocorallia). Mol. Ecol. 13:1495–1505 McNeill DF, Budd AF, Borne FP. 1997. An earlier (late Pliocene) first appearance of the reef-building coral Acropora palmata: stratigraphic and evolutionary implications. Geology 25:891–94 Metz EC, Palumbi SR. 1996. Positive selection and sequence rearrangements generate extensive polymorphism in the gamete recognition protein bindin. Mol. Biol. Evol. 13:397–406 Miller DJ, van Oppen MJH. 2003. A “fair go” for coral hybridization. Mol. Ecol. 12:805–7 Miller KJ, Ayre DJ. 2004. The role of sexual and asexual reproduction in structuring high latitude populations of the reef coral Pocillopora damicornis. Heredity 92:557– 68 Miller KJ, Babcock RC. 1997. Conflicting morphological and reproductive species boundaries in the coral genus Platygyra Biol. Bull. 192:98–110 Miller KJ, Benzie JAH. 1997. No clear genetic distinction between morphological species within the coral genus Platygyra Bull. Mar. Sci. 61:907–17 Modrich P, Lahue R. 1996. Mismatch repair in replication fidelity, genetic recombi- nation and cancer biology. Annu. Rev. Biochem. 65:101–33 Naisbit RE, Jiggins CD, Mallet J. 2003. Mimicry: developmental genes that contribute to speciation. Evol. Dev. 5:269–80 Odorico DM, Miller DJ. 1997. Variation in the ribosomal internal transcribed spacers and 5.8S rDNA among five species of Acropora (Cnidaria; Scleractinia): patterns of variation consistent with reticulate evolution. Mol. Biol. Evol. 14:465–73 Otto SP, Whitton J. 2000. Polyploid incidence and evolution. Annu. Rev. Genet. 34:401–37 Palumbi SR. 1994. Genetic divergence, reproductive isolation, and marine speciation.

by NORTHEASTERN UNIVERSITY on 01/11/07. For personal use only. Annu. Rev. Ecol. Syst. 25:547–72 Potts BM, Reid JB. 1988. Hybridization as a dispersal mechanism. Evolution 42:1245– 55 Annu. Rev. Ecol. Evol. Syst. 2006.37:489-517. Downloaded from arjournals.annualreviews.org Ramsey J, Schemske DW. 1998. Pathways, mechanisms, and rates of polyploid formation in flowering plants. Annu. Rev. Ecol. Syst. 29:467–501 Richmond RH, Hunter CL. 1990. Reproduction and recruitment of corals: compar- isons among the Caribbean, the Tropical Pacific, and the Red Sea. Mar. Ecol. Prog. Ser. 60:185–203 Rieseberg LH. 1997. Hybrid origins of plant species. Annu. Rev. Ecol. Syst. 28:359–89 Rieseberg LH, Raymond O, Rosenthal DM, Lai Z, Livingstone K, et al. 2003. Ma- jor ecological transitions in wild sunflowers facilitated by hybridization. Science 301(5637):1211–16

www.annualreviews.org • Hybridization in Corals 515 ANRV292-ES37-18 ARI 17 October 2006 7:29

Rosenfield JA, Kodric-Brown A. 2003. Sexual selection promotes hybridization between Pecos pupfish, Cyprinodon pecosensis, and sheepshead minnow, C-variegatus. J. Evol. Biol. 16:595–606 Rutzler¨ K, Macintyre IG. 1982. Habitat distribution and community structure of the barrier reef complex at Carrie Bow Cay, Belize. In The Atlantic Barrier Reef Ecosystem at Carrie Bow Bay, Belize, I. Structure and Communities,KRutzler,¨ IG Macintyre, eds., pp. 9–45. Washington: Smithsonian Inst. Press. 539 pp. Salzburger W, Baric S, Sturmbauer C. 2002. Speciation via introgressive hybridization in east African cichlids? Mol. Ecol. 11:619–25 Seehausen O. 2004. Hybridization and adaptive radiation. Trends Ecol. Evol. 19:198– 207 Shaw KL. 2002. Conflict between nuclear and mitochondrial DNA phylogenies of a recent species radiation: what mtDNA reveals and conceals about modes of speciation in Hawaiian crickets. Proc. Natl. Acad. Sci. USA 99:16122–27 Simpson CJ. 1991. Mass spawning of corals on Western Australian reefs and com- parisons with the Great Barrier Reef. J. R. Soc. West. Aust. 74:85–91 Soltis DE, Soltis PS. 1999. Polyploidy: recurrent formation and genome evolution. Trends Ecol. Evol. 14:348–52 Spolsky CM, Phillips CA, Uzzell T. 1992. Antiquity of clonal salamander lineages revealed by mitochondrial DNA. Nature 356:706–10 Stebbins GL. 1959. The role of hybridization in evolution. Proc. Am. Philos. Soc. 103:231–51 Szmant AM. 1986. Reproductive ecology of Caribbean reef corals. Coral Reefs 5:43–54 Szmant AM, Weil E, Miller MW, Colon DE. 1997. Hybridization within the species complex of Montastrea annularis. Mar. Biol. 129:561–72 Templeton AR. 1989. In Speciation and its Consequences, ed. D Otte, JA Endler. New York: Sinauer. 670 pp. TompkinsDM, Mitchell RA, Bryant DM. 2006. Hybridization increases measures of innate and cell-mediated immunity in an endangered bird species. J. Anim. Ecol. 75:559–64 Ungerer MC, Baird SJE, Pan J, Rieisberg LH. 1998. Rapid hybrid speciation in wild sunflowers. Proc. Natl. Acad. Sci. USA 95:11757–762 Vacquier VD, Corner KR, Stout CD. 1990. Species specific sequences of abalone by NORTHEASTERN UNIVERSITY on 01/11/07. For personal use only. lysine, the sperm protein that creates a hole in the egg envelope. Proc. Natl. Acad. Sci. USA 87:5792–96 van Herwerden L, Choat JH, Dudgeon CD, Carlos G, Newman SJ, et al. 2006. Annu. Rev. Ecol. Evol. Syst. 2006.37:489-517. Downloaded from arjournals.annualreviews.org Contrasting patterns of genetic structure in two species of the coral trout Plec- tropomus (Serranidae) from east and west Australia: introgressive hybridization or ancestral polymorphisms. Mol. Phyl. Evol. In press van Oppen MJH, McDonald BJ, Willis BL, Miller DJ. 2001. The evolutionary history of the coral genus Acropora (Scleractinia, Cnidaria) based on a mitochondrial and a nuclear marker: Reticulation, incomplete lineage sorting or morphological convergence? Mol. Biol. Evol. 18:1315–29 van Oppen MJH, Willis BL, van Rheede T, Miller DJ. 2002. Spawning times, reproductive compatibilities and genetic structuring in the Acropora aspera group:

516 Willis et al. ANRV292-ES37-18 ARI 17 October 2006 7:29

evidence for natural hybridization and semipermeable species boundaries in corals. Mol. Ecol. 11:1363–76 van Oppen MJH, Willis BL, van Vugt H, Miller DJM. 2000. Examination of species boundaries in the Acropora cervicornis group (Scleractinia, Cnidaria) using nuclear DNA sequence analysis. Mol. Ecol. 9:1363–73 Vaughan TW. 1901. The stony corals of the Porto Rican waters. Bull. US Fish Com- miss. 1900 20:290–320 Veron JEN. 1995. Corals in Space and Time. The Biogeography and Evolution of the Scleractinia. Sydney: UNSW Press. 321 pp. Veron JEN. 2000. Corals of the World. Vol. 1. Townsville, Aust.: AIMS. 463 pp. Vollmer SV, Palumbi SR. 2002. Hybridization and the evolution of reef coral diversity. Science 296:2023–25 Vollmer SV, Palumbi SR. 2004. Testing the utility of internally transcribed spacer sequences in coral phylogenetics. Mol. Ecol. 13:2763–72 Vrijenhoek RC. 1984. Ecological differentiation among clones: The frozen niche variation model. In Population Biology and Evolution, ed. K Wohrmann, V Loeschecke. Berlin: Springer-Verlag Wallace CC. 1999. Staghorn Corals of the World: A Revision of the Genus Acropora. Collingwood, Aust.: CSIRO Publishing. 421 pp. 1st ed. Wallace CC, Willis BL. 1994. The systematics of Acropora: The effect of new biological findings on species concepts. Annu. Rev. Ecol. Syst. 25:237–62 Wells JW. 1973. New and old scleractinian corals from Jamaica. Bull. Mar. Sci. 23:16– 58 White MJ. 1978. Modes of Speciation. San Francisco: Freeman Whittemore AT, Schaal BA. 1991. Interspecific gene flow in sympatric oaks. Proc. Natl. Acad. Sci. USA 88:2540–44 Willis BL, Babcock RC, Harrison PL, Oliver JK, Wallace CC. 1985. Patterns in the mass spawning of corals on the Great Barrier Reef. Proc. Fifth Int. Coral Reef Symp. Tahiti 4:343–48 Willis BL, Babcock R, Harrison P, Wallace C. 1997. Experimental hybridization and breeding incompatibilities within the mating systems of mass spawning corals. Coral Reefs 16:S53–65 Wolstenholme JK. 2004. Temporalreproductive isolation and gametic compatibility by NORTHEASTERN UNIVERSITY on 01/11/07. For personal use only. are evolutionary mechanisms in the Acropora humilis species group (Cnidaria; Scleractinia). Mar. Biol. 144:567–82 Wolstenholme JK, Wallace CC, Chen CA. 2003. Species boundaries within the Annu. Rev. Ecol. Evol. Syst. 2006.37:489-517. Downloaded from arjournals.annualreviews.org Acropora humilis species group (Cnidaria; Scleractinia): a morphological and molecular interpretation of evolution. Coral Reefs 22:155–66

www.annualreviews.org • Hybridization in Corals 517 Contents ARI 20 September 2006 18:4

Annual Review of Ecology, Evolution, Contents and Systematics Volume 37, 2006

Birth-Death Models in Macroevolution Sean Nee pppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppp1 The Posterior and the Prior in Bayesian Phylogenetics Michael E. Alfaro and Mark T. Holder pppppppppppppppppppppppppppppppppppppppppppppppppppppp19 Unifying and Testing Models of Sexual Selection Hanna Kokko, Michael D. Jennions, and Robert Brooks pppppppppppppppppppppppppppppppppppp43 Genetic Polymorphism in Heterogeneous Environments: The Age of Genomics Philip W. Hedrick ppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppp67 Ecological Effects of Invasive Arthropod Generalist Predators William E. Snyder and Edward W. Evans ppppppppppppppppppppppppppppppppppppppppppppppppppp95 The Evolution of Genetic Architecture Thomas F. Hansen ppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppp123 The Major Histocompatibility Complex, Sexual Selection, and Mate Choice Manfred Milinski pppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppp159 Some Evolutionary Consequences of Being a Tree Rémy J. Petit and Arndt Hampe ppppppppppppppppppppppppppppppppppppppppppppppppppppppppppp187 by NORTHEASTERN UNIVERSITY on 01/11/07. For personal use only. Late Quaternary Extinctions: State of the Debate Paul L. Koch and Anthony D. Barnosky ppppppppppppppppppppppppppppppppppppppppppppppppppp215 Annu. Rev. Ecol. Evol. Syst. 2006.37:489-517. Downloaded from arjournals.annualreviews.org Innate Immunity, Environmental Drivers, and Disease Ecology of Marine and Freshwater Invertebrates Laura D. Mydlarz, Laura E. Jones, and C. Drew Harvell ppppppppppppppppppppppppppppppp251 Experimental Methods for Measuring Gene Interactions Jeffery P. Demuth and Michael J. Wade ppppppppppppppppppppppppppppppppppppppppppppppppppp289 Corridors for Conservation: Integrating Pattern and Process Cheryl-Lesley B. Chetkiewicz, Colleen Cassady St. Clair, and Mark S. Boyce ppppppppppp317

v Contents ARI 20 September 2006 18:4

The Population Biology of Large Brown Seaweeds: Ecological Consequences of Multiphase Life Histories in Dynamic Coastal Environments David R. Schiel and Michael S. Foster ppppppppppppppppppppppppppppppppppppppppppppppppppppp343 Living on the Edge of Two Changing Worlds: Forecasting the Responses of Rocky Intertidal Ecosystems to Climate Change Brian Helmuth, Nova Mieszkowska, Pippa Moore, and Stephen J. Hawkins ppppppppppp373 Has Vicariance or Dispersal Been the Predominant Biogeographic Force in Madagascar? Only Time Will Tell Anne D. Yoder and Michael D. Nowak pppppppppppppppppppppppppppppppppppppppppppppppppppp405 Limits to the Adaptive Potential of Small Populations Yvonne Willi, Josh Van Buskirk, and Ary A. Hoffmann pppppppppppppppppppppppppppppppppp433 Resource Exchange in the Rhizosphere: Molecular Tools and the Microbial Perspective Zoe G. Cardon and Daniel J. Gage pppppppppppppppppppppppppppppppppppppppppppppppppppppppp459 The Role of Hybridization in the Evolution of Reef Corals Bette L. Willis, Madeleine J.H. van Oppen, David J. Miller, Steve V. Vollmer, and David J. Ayre ppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppp489 The New Bioinformatics: Integrating Ecological Data from the Gene to the Biosphere Matthew B. Jones, Mark P. Schildhauer, O.J. Reichman, and Shawn Bowers pppppppppp519 Incorporating Molecular Evolution into Phylogenetic Analysis, and a New Compilation of Conserved Polymerase Chain Reaction Primers for Animal Mitochondrial DNA Chris Simon, Thomas R. Buckley, Francesco Frati, James B. Stewart, and Andrew T. Beckenbach pppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppp545 The Developmental, Physiological, Neural, and Genetical Causes and

by NORTHEASTERN UNIVERSITY on 01/11/07. For personal use only. Consequences of Frequency-Dependent Selection in the Wild Barry Sinervo and Ryan Calsbeek pppppppppppppppppppppppppppppppppppppppppppppppppppppppppp581 Carbon-Nitrogen Interactions in Terrestrial Ecosystems in Response Annu. Rev. Ecol. Evol. Syst. 2006.37:489-517. Downloaded from arjournals.annualreviews.org to Rising Atmospheric Carbon Dioxide Peter B. Reich, Bruce A. Hungate, and Yiqi Luo pppppppppppppppppppppppppppppppppppppppppp611 Ecological and Evolutionary Responses to Recent Climate Change Camille Parmesan ppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppp637

Indexes

Cumulative Index of Contributing Authors, Volumes 33–37 ppppppppppppppppppppppppppp671 Cumulative Index of Chapter Titles, Volumes 33–37 pppppppppppppppppppppppppppppppppppp674

vi Contents